INDEX

Explanations

references to numeric values and identifiers in a structured format

oai_token-act-pair · gpt-4o-mini Triggered by @bot

conditions and comparisons

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

IntoConstraints

-0.64

AddTagHelper

-0.63

 дописавши

-0.58

 CreateTagHelper

-0.55

 ExecuteAsync

-0.53

 '\\;'

-0.53

Hentet

-0.52

دانشنامهٔ

-0.52

">//

-0.50

MessageOf

-0.50

POSITIVE LOGITS

SourceChecksum

0.39

 implies

0.39

且

0.38

 excluding

0.38

Geografi

0.37

ylabel

0.37

AllowAnonymous

0.36

getActive

0.35

 excludes

0.35

 Superhosts

0.35

Activations Density 0.014%

references to numeric values and identifiers in a structured format

conditions and comparisons

No Comments

No Known Activations